added H quicklook code from maven_iuvs_chaffin

[planetarymike]

This is an initial commit of some very messy (but working!) code that generates H Lyman alpha quicklooks. Much to improve before this can be merged, I think.

Pull Request Checklist

• Code Follows PEP8 guidelines (explain any exceptions)
• New routines don't duplicate existing functionality
• [n/a] Original authors of modified routines have been added as reviewers

[planetarymike]

note that some care is needed to get this working on your local system--- Mike with consult with Eryn on path forward.

[planetarymike]

Here is my giant list of to-dos from the notebook where I originally developed this code:

---

to-do

Continue tweaks to fit_line:

• improve flatfield
• improve MUV subtraction of apoapsis --- sonal knows how to do this

Plotting tweaks:

• deal with 24 periapse obs cases like on orbit 1694
  • higher numbers of periapse files should be from nadir-looking aurora campaigns:
    (per Justin: "A check for this type of nadir viewing auroral scan is that it should have 12 spatial bins and 3 integrations per file. There can be regular limb scans in other parts of the periapse segment.")
    ^^^^early versions of this sequence have 7 spatial bins
  • there are indeed some files with the usual number of ints/bins --- need to display only these on the viz
• Ancillary plot stuff
  • add a colorbar for MCS dust
  • add more labels to ancillary plots
  • add tau=1 and line center brightness (Ly a) to EUVM plot (probably requires interpolation)
• width of swaths on apoapse orbits > 11000:
  • MRH point is on near side of planet even though observation extends past limb. Draw these at location of limb instead.
• Check printed altitudes on limb plots are using the right y-coordinate (may be backward on inbound segment)
• Units on EUVM
• horizontal alignment of periapsis swaths: make sure this is aligned with along-orbit direction. Right now it's a fixed based on spatial binning

File tweaks:

• remove old/bad files from stage that are on the wrong orbit:
  • wtf is going on with orbits like 8240? Looks like every observation is doubled or tripled
    • this happens for orbits 8235-8250
  • due to a CenturyLink outage in early 2019. Bad files are stage tagged and can be removed
• write some checker routine to be run in sync_latest_files
  • did this and I have some stage files that were never made by production and are also not on the PDS
  • sent a list of these to Sonal (September 2020 sometime?)

Bugs/corner cases:

• orbit 855: serious issues with apoapse images, likely due to line integration routine
• orbit 333: HV turns off before observation ends, even though file reports it on.
  • What to do? implement total counts per image filter? Average counts < 0?
• file mvn_iuv_l1b_inspace-orbit00341-ech_20141202T034550_v13_r01.fits.gz seems to be assigned the wrong orbit, orbit 341 already has muv and fuv obs. Orbit 340 has an echelle obs already?
  • This appears to happen many times in the orbit range 0-1000
• some apoapse files use nonlinear binning. Ex: orbit 386
• issues with parsing integrated report file on orbit 972 and others
  • Possibly due to safe mode entry?
• orbit 110 apoapsis: "ValueError: x and y arguments to pcolormesh cannot have non-finite values or be of type numpy.ma.core.MaskedArray with masked values"
  • this is because of NaN values in PIXEL_VEC --- why are these present???
  • make a list of orbits where these occur, see if there are any recent examples
• NaNs in periapse / other geometry
  • Same issue as above for apoapsis, likely there is a bug in the mirror kernel
  • sometimes the whole map is gray? Gonna guess this is related to the above
• v03 files are present intermittently in the data--- do these contain the same data as later version files?
• check date alignment of ancillary values--- EUVM? these look to not be centered on the orbit
• MAYBE instead of not making a file, print a blank canvas with the orbit and a warning that no files were found
  • that way if there's no file there was definitely an error

Wrap up and package integrated report/filename comparator

Code to select only orbits where certain subsets are populated (apoapse, side segment, etc.)

---

shortly after that

Timeline plots in new right panel:

• depends on ly-a brightness being saved somewhere
• timelines of periapsis/side segment brightness
• something similar for apoapsis?

---

longer term

• print special event text if stellar occ, deep dip, conjunction, caveats file, etc.
  ** need time ranges for these events --- could get from Ops calendar
  ** there are also several lists in the caveats file

Handle corner cases

• convert generic raise to specific python error
• handle bad pointing for apoapsis
  • like in orbit 3498
  • not clear that this can be fixed

Lower priority tweaks:

• map stuff
  • add label for peri+side segment scans
  • decide how to show+label side segment scans
  • add better+fast limb detection for apoapsis swath boundaries
  • label apoapsis boundaries
• store filenames in metadata only and use the extra space for something else?

More recent ancillary datasets:

• MCS especially--- any way to get this from the PDS?
• do SWIA/EUVM processing locally?

Add generic hook for an input function to generate plot intensities for each observation

• allows others to use the code to make plots of their own